A Novel Multi-Objective Project Portfolio Selection and Scheduling Problem

A common problem faced by organisations is how to most effectively select and schedule a portfolio of projects subject to a limited budget. This problem is referred to as the project portfolio selection and scheduling problem (PPSSP). In the context of the Australian Department of Defence (DoD), the PPSSP arises as part of the future force design (FFD) process. Recently, a novel PPSSP formulation was shown to demonstrate the characteristics required to address the modern FFD problem. However, the FFD process is inherently multi-objective while the aforementioned problem formulation was limited to only maximisation of the total portfolio value. This study proposes a bi-objective extension of this model and provides an empirical analysis of three multi-objective meta-heuristics to address 81 instances of the proposed problem. Further insights about the solutions, used to aid the decision maker when selecting a portfolio for implementation, are provided using knee-point analysis and scalarisation.